A founder or CEO reaches out: "I think we need a COO. Can you help us? Or do you know someone?"
My answer is: "No — I don't do that full-time anymore, and you may not actually need one. So before we dive into solutions, let's figure out what your actual problem is. I can help you with that. Let's go through your ops."
The moment things really break
When your startup grows — in people, revenue, or investors — what you've been doing operationally will either ask you to step up to the next level, or it will break. And then it will distract you from where your superpowers should be applied. Or it will burn you out. Or it will block your team. Once you get close to that point, ask help! Other experienced founders or operators are often happy to help guide you through this phase. The founders who reach out to me with those growing pains are often in one of two situations:
One: The all-rounder founder who hit the ceiling. They did everything themselves for a while, and they were good at it — better than the engineers, designers, or go-to-market people they'd otherwise hand it to. Around 15 people, it stops being a 10% thing on the side. You're the one still writing the offer letters, making the financial report, chasing late paying customers and ordering laptops for new hires, and setting next quarter's strategy. They're burning out, dropping operational balls, or dropping things that are the core of their job. Often all of the above.
Two: The visionary product founder who's not into ops. They duct-taped it through the early phases. It wasn't great, but it wasn't catastrophic. Now things are seriously broken. Your week is back-to-back product and customer calls, and the ops 'system' is a Notion not updated in months and the team asking 'how do I even get the new hire a laptop and Slack?' and Angel investors asking 'when will you send a report, it has been months'. Some founders patch it with more duct tape and get an admin person to handle some of it, or — if they were a bit smarter about it — a chief of staff. Sometimes that still breaks, and that's when they reach out.
The good news: this is fixable, and you can start this week
The diagnostic framework helps you figure out how to fix it: we go through your processes, one by one, and look at the state each one is in. One call or self diagnosis gets you a map of what's actually broken versus what just feels that way. The patterns in your strengths and weaknesses. And where to focus next.
Here's what that usually looks like for each of the two situations.
For the all-rounder founder who hit the ceiling, the pattern is usually this: a lot of processes are still owned by you, undocumented, they exist only in your head. That sounds bad, but it's actually the easier diagnosis — because the fix has an obvious first step. Pick one process. Record a voice note on your walk home and have Claude summarise it into a one-pager. Or do a quick Loom while you're doing the thing. Share it with someone, and let them try it next time. Lightweight documentation, then one person other than you who can do it. Repeat for the next process. Most of what feels overwhelming becomes manageable once that motion starts.
For the visionary founder who duct-taped it through, the picture is more mixed: some things are genuinely handed over, some are still broken, and there's often a chief of staff or admin person quietly drowning in the parts you don't see. The issue is no longer that no one knows how it works, or that they don't work, but the one doing it is overwhelmed. The diagnostic gives founder and ops person a structured overview of what's actually broken versus what just feels overwhelming but is fine for you phase. Quite often it's not as bad as the chief of staff fears, just a bit worse than the founder thinks ;-). Knowing which is which is half the work and will be the result of going through the framework.
In both cases, getting help in creating the overview of your ops is the first step to unblock your ops worries.
How the framework came about
Last year I helped a handful of founders in exactly these situations. At first I improvised my analysis. After a few cases, I noticed I was doing roughly the same thing each time — and with some long conversations with Claude I worked out what the method actually was. I turned it into a structured questionnaire. Ruben then challenged me to make it into a web app. And there was the diagnostic framework (shout out to Ruben for always challenging me!).
How it works
The framework is a set of questions per process that determines where you stand. Most ops frameworks tell you what good looks like in general. This one tells you what good looks like at your specific stage, and where you are right now relative to that. Which processes apply depends on what stage you're at as a company:
- Three founders and a dog? You need almost nothing.
- 35 people, paying customers, investors on the cap table? You need a real operational backbone.
For each process we look at how broken it is right now, who owns it (and whether it has a clear owner), whether it's documented, what tooling supports it, and how much extra scale it can handle.
Why this exists
My ambition is to make this kind of operational clarity available to more founders and organisations than I could ever personally help. Through the framework being open and free, through other ops people picking it up and using it with their own clients, and increasingly with AI doing more of the heavy lifting.
A lot of my work is pro bono with non-profits and mission-driven organisations who can't easily afford a fractional COO but need this thinking just as much. The framework being open is what makes that scalable. I keep doing some commercial work too — partly to fund my work for non-profits, partly because working with scale-ups and growth-stage startups is what keeps the framework honest in those settings. The two feed each other.
Try it, use it — it's open source. You can self-diagnose, but it works better with an experienced ops person walking you through it. They'll know what 'good' looks like for your stage, and can review your actual processes. Ask someone in your network, or reach out — we'll figure out whether you need a COO, or just AI-augmented duct tape.......